Boosting Product Recommendations: Simplifying the Process for Wellness Experts at Wrkout
As the lead UX/UI designer at Wrkout, a startup aimed at improving health outcomes for wellness professionals and their clients, my mission was to enhance the Wrkout Store app. Our app facilitates wellness professionals in recommending health and fitness products to their clients, with professionals receiving cash rewards for client purchases. Over the course of a 3-week project, I was tasked with improving the number of recommendations shared through app.
UX/UI Designer
Product Manager, FE Dev
iOS / Android App
Figma
3 weeks
Role
Team
Platform
Tools
Duration
The Goal
Our goal was to boost the number of product recommendations sent through the platform. With Wrkout's revenue dependent on users purchasing recommended products, increasing these recommendations was crucial. We aimed for changes that were low effort to develop, allowing us to quickly gauge the success of the update.
The Problem
Through a series of qualitative user interviews, we uncovered two key issues hindering recommendation creation.
Firstly, we found that the recommendation flow had unnecessary steps, causing friction and offering little value to users. For instance, few users utilized the 'share by email' feature, adding an extra decision point.
Secondly, trainers tended to recommend only familiar products and brands. This signalled that users may not be finding the products and brands they trust, therefore reducing potential recommendation opportunities.
Simplifying the Recommendation Flow
During user interviews, we discovered that most trainers preferred creating share links over email when recommending products. Eliminating the share by email option offered several benefits: users could recommend products more easily without entering client emails, and the code base could be simplified with automated recommendation emails.
I also proposed that we place the ‘product gathering’ and ‘add message’ steps onto two distinct screens. This would allow users to focus on a single task per page, therefore reducing the cognitive load.
Additionally, we made the message optional to reduce friction further. During interviews it was discovered that trainers often provided recommendations during real-time conversations with clients, where a written message was unnecessary and hindered their ability to share quick recommendations in the moment.
Increasing Product Discoverability
We received feedback indicating that trainers felt comfortable recommending only products they knew and trusted. To increase recommendations, it was crucial to ensure that trainers could easily discover these products.
Drawing on the assumption that trainers have strong brand allegiances, we believed that showcasing the brands carried in the Wrkout Store would provide a quick overview of the product catalog and lead to more recommendations. To achieve this, we decided to focus on a brand-first experience when users landed on the home page. This involved implementing a 'featured brand' hero carousel followed by a section leading to our 'all brands' page.
Impact / Reflections
The final updates to the Home Screen and recommendation flow received approval from the product team and key stakeholders. I collaborated closely with the development team to create developer-ready versions of the mocks, ensuring seamless implementation of design decisions. Following a successful launch, the update was rolled out to all users. The success of this initiative was measured by tracking the number of recommendations shared over a period, which showed a small increase.
Reflecting on this update, there are a few limitations I would approach differently in the future. At the time, due to resource constraints as a startup, we unfortunately lacked the finer analytics to track click-through rates on the home page and recommendation flows. Our plan was to collect more data in these areas to further fine-tune our improvements. While I relied largely on qualitative user feedback to inform my design decisions, time constraints prevented thorough user testing before development. Given more time, I would have tested the home page for ‘first click’ tendencies to gauge new users' interest in discovering new brands. Additionally, I would have tested the proposed recommendation flow for ‘task success rate’, ‘user error rate’, and ‘time on task’ to gain insight into any unforeseen shortcomings with the flow.